> ## Documentation Index
> Fetch the complete documentation index at: https://docs.baseten.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Libraries and tools

> Deploy models, run inference, and manage training jobs with Baseten SDKs and tools.

## Command-line tools

<CardGroup cols={2}>
  <Card title="Baseten CLI" icon="terminal">
    Deploy and manage models from your terminal. Install with Homebrew on macOS or Linux:

    ```sh theme={"system"}
    brew tap basetenlabs/baseten
    brew trust --formula basetenlabs/baseten/baseten
    brew install baseten
    ```

    [GitHub](https://github.com/basetenlabs/baseten-cli) · [Docs](/reference/cli/baseten/overview)
  </Card>

  <Card title="Truss CLI" icon="terminal">
    Package models and run Chains, Loops, and Training commands.

    ```sh theme={"system"}
    pip install truss
    ```

    [GitHub](https://github.com/basetenlabs/truss) · [Docs](/reference/cli/truss/overview)
  </Card>

  <Card title="Baseten Switch" icon="shuffle">
    Configure coding agents to use Model APIs on macOS. Baseten Switch is in beta.

    ```sh theme={"system"}
    brew trust basetenlabs/baseten
    brew install basetenlabs/baseten/baseten-switch
    ```

    [Docs](/reference/cli/baseten-switch)
  </Card>
</CardGroup>

## Management SDKs

Call the [Management API](/reference/management-api/overview) from your application.

<CardGroup cols={2}>
  <Card title="Python" icon="python" iconType="brands">
    Manage models and deployments with synchronous or asynchronous Python methods.

    ```sh theme={"system"}
    pip install baseten
    ```

    [GitHub](https://github.com/basetenlabs/baseten-python) · [Docs](https://basetenlabs.github.io/baseten-python/api/baseten.client.html)
  </Card>

  <Card title="TypeScript and JavaScript" icon="js" iconType="brands">
    Manage models and deployments from Node.js applications.

    ```sh theme={"system"}
    npm install @basetenlabs/client
    ```

    [GitHub](https://github.com/basetenlabs/baseten-js) · [Docs](https://basetenlabs.github.io/baseten-js/client/)
  </Card>

  <Card title="Go" icon="golang" iconType="brands">
    Manage models and deployments from Go applications.

    ```sh theme={"system"}
    go get github.com/basetenlabs/baseten-go
    ```

    [GitHub](https://github.com/basetenlabs/baseten-go) · [Docs](https://pkg.go.dev/github.com/basetenlabs/baseten-go/client)
  </Card>
</CardGroup>

## Model development and training

<CardGroup cols={2}>
  <Card title="Truss SDK" icon="box">
    Package and deploy models from Python. Import the package as `truss`.

    ```sh theme={"system"}
    pip install truss
    ```

    [GitHub](https://github.com/basetenlabs/truss) · [Docs](/reference/sdk/truss/overview)
  </Card>

  <Card title="Chains SDK" icon="link">
    Define chainlets and compose inference pipelines. Import the package as `truss_chains`.

    ```sh theme={"system"}
    pip install truss
    ```

    [GitHub](https://github.com/basetenlabs/truss) · [Docs](/reference/sdk/chains)
  </Card>

  <Card title="Loops SDK" icon="infinity">
    Run LoRA training and generate samples with Python 3.12 or later. Import the package as `baseten.loops`.

    ```sh theme={"system"}
    pip install baseten-loops
    ```

    [PyPI](https://pypi.org/project/baseten-loops/) · [Docs](/reference/sdk/loops/overview)
  </Card>

  <Card title="Training SDK" icon="dumbbell">
    Configure compute, runtime, and storage for training jobs. Import the package as `truss_train`.

    ```sh theme={"system"}
    pip install truss
    ```

    [GitHub](https://github.com/basetenlabs/truss) · [Docs](/reference/sdk/training)
  </Card>
</CardGroup>

## Inference clients

The [Performance Client](/reference/sdk/performance-client/overview) sends concurrent embedding, reranking, classification, and batch requests to model endpoints.

<CardGroup cols={2}>
  <Card title="Python Performance Client" icon="python" iconType="brands">
    Send concurrent inference requests with synchronous or asynchronous Python methods.

    ```sh theme={"system"}
    pip install baseten_performance_client
    ```

    [GitHub](https://github.com/basetenlabs/truss/tree/main/baseten-performance-client/python_bindings) · [Docs](/reference/sdk/performance-client/python)
  </Card>

  <Card title="Node.js Performance Client" icon="js" iconType="brands">
    Send concurrent inference requests from Node.js applications.

    ```sh theme={"system"}
    npm install @basetenlabs/performance-client
    ```

    [GitHub](https://github.com/basetenlabs/truss/tree/main/baseten-performance-client/node_bindings) · [Docs](/reference/sdk/performance-client/node)
  </Card>

  <Card title="Rust Performance Client" icon="rust" iconType="brands">
    Send concurrent inference requests with Tokio.

    ```sh theme={"system"}
    cargo add baseten_performance_client_core
    ```

    [GitHub](https://github.com/basetenlabs/truss/tree/main/baseten-performance-client/core) · [Docs](/reference/sdk/performance-client/rust)
  </Card>
</CardGroup>

See [Batching and concurrency](/reference/sdk/performance-client/overview#batching-and-concurrency) to control batch size and concurrent requests.

## Terraform

<CardGroup cols={2}>
  <Card title="Terraform provider" icon="cubes">
    Manage Baseten resources with Terraform. Add the provider to your configuration:

    ```hcl theme={"system"}
    terraform {
      required_providers {
        baseten = {
          source  = "basetenlabs/baseten"
          version = "~> 0.1"
        }
      }
    }
    ```

    Run `terraform init` to install the provider.

    [GitHub](https://github.com/basetenlabs/terraform-provider-baseten) · [Docs](https://registry.terraform.io/providers/basetenlabs/baseten/latest/docs)
  </Card>
</CardGroup>

## Related references

* [CLI reference](/reference/cli/index) for terminal commands and scripts.
* [API reference](/reference/overview#api-reference) for HTTP endpoints.
* [Configuration reference](/reference/configuration) for Truss and engine settings.